Xi Liang
Senior Software Engineer at Databricks · Ph.D. in Computer Science, University of Chicago.
News
In March 2022, I successfully defended my dissertation titled “Synopses for Efficient and Reliable Approximate Query Processing” and I joined Databricks as a software engineer.
I appreciate all the help and support I get from my advisor (Prof. Sanjay Krishnan), faculties and collaborators in the ChiData group (Prof. Mike Franklin, Prof. Aaron Elmore, Prof. Raul Castro Fernandez, Dr. Stavros Sintos, Dr. Zechao Shang), and many friends and peers in the CS department.
About Me
I earned my Ph.D. in Computer Science advised by Prof. Sanjay Krishnan in the ChiData group at the University of Chicago. My research focused on designing and building efficient data systems with a special interest in approximate query processing.
Before joining the department I completed my M.S. in Computer Science under the supervision of Prof. I-Chen Wu at National Chiao Tung University (NCTU, Hsinchu, Taiwan). Prior to that, I worked as a software engineer after I graduated from Harbin Institute of Technology (HIT, Harbin, China) where I majored in Software Engineering.
I enjoy distance running especially on the beautiful Chicago Lakefront Trail. I finished three full marathons so far: Taipei Marathon in 2014, Hsinchu City Marathon in 2015, and Chicago Marathon in 2018. I hope I can finish all the World Marathon Majors someday.
Selected Publications
JanusAQP: Efficient Partition Tree Maintenance for Dynamic Approximate Query Processing
- Xi Liang, Stavros Sintos, Sanjay Krishnan. ICDE 2023.
- Paper
Combining Aggregation and Sampling (Nearly) Optimally for Approximate Query Processing
- Xi Liang, Stavros Sintos, Zechao Shang, Sanjay Krishnan. SIGMOD 2021.
- Paper
Fast and Reliable Missing Data Contingency Analysis with Predicate-Constraints
Opportunistic View Materialization with Deep Reinforcement Learning
Patents
Compile-time processing of extract, transform, load process
- Michael Armbrust, Yannis Papakonstantinou, …, Xi Liang, et al. US Patent 12,517,905 (Databricks), granted Jan 2026.
Incremental execution of extract, transform, load process using microtechniques architecture
- Michael Armbrust, Yannis Papakonstantinou, …, Xi Liang, et al. US Patent 12,386,833 (Databricks), granted Aug 2025.
